As small tribes farmed the Dorset landscape of the Iron Age, the political environment in the Roman Emipire was changing. Determined to expand they set their sights on Albion, forcing their way through with armed legions on newly developed roads, and renamed the country Britannia. New technologies, building practices, tools, skills, materials and beliefs were merged into existing life, most likely through awe and admiration rather than battle. The landscape changed dramatically, becoming the framework of the future, but it was not to last. Today their mark is still clearly visible in present day infrastructure, earthworks, foundations, fields and names with many treasures often unearthed.
Iron Age
Dorset, along with parts of south Somerset, Wiltshire and the New Forest, was inhabited by the Durotriges tribe; the name given to them by the Romans. There are a number of theories regarding the origin of the name, the most widely known being the Latinised version of the Common Brittonic word ‘Duro’, possibly translating to a ‘place with fist-sized pebbles’, referring to Chesil Beach. Alternatively the ‘Dur’ could be a Celtic word translating to land or water dwellers, in reference to the farmed valley of the River Frome. Although hillforts are regarded as abandoned in the late Iron Age period, the hillforts of Dorset were still utilised, some occupied while others possibly used for ceremonial or community purposes. Meanwhile the majority of the population had moved down into the, freshly deforested, fertile valleys, creating a loosely knit confederation of smaller tribal groups.
The Durotriges had their own coinage and the trade in pottery and grain was strong. The natural harbour of Christchurch (soon to become Dunium) provided easy access from the inland rivers of the Stour and Avon across the channel to Gaul and the Roman Empire. However, during the reign of Julius Caesar, trade soon dried up as Roman politics led to the realignment of priorities to deal with only those who were believed to be allies. Prior to the Roman invasion the population was high, land had become over farmed, and trade was at an all-time low. Skeletons unearthed from the period show signs of malnutrition suggesting that the population was also weak.
Hengistbury Head (Port Dunium)
Christchurch, originally known as Twynham, sits on the confluence of the rivers Avon and Stour. The name Twynham derives from the old English tweoxn (between) and ēam (stream), translating to ‘between the streams’. By 100BC, the late Iron Age, Christchurch Harbour had become a substantial international port exporting and importing goods such as iron, silver, bronze and even Italian wine. This perfectly placed landscape, providing evidence of settlement, activity and trade, leads to the theory that it could have been the first ever urban settlement in England. When the Romans arrived, little changed as trade was already part of society. The harbour was working efficiently, with the help of the natural landscape, and so the Romans used it just as equally, but in the company of more exotic goods. A Roman ship was discovered in the harbour in 1910. The person who discovered it sent a few artefacts to The British Museum, but died in World War I, before he could show anyone where it was!
Maiden Castle
Maiden Castle is one of the largest hillforts in Europe. An excavation in the 1930’s around the eastern entrance’s ramparts, produced numerous skeletons, one with an axe embedded into a spine. The finds are now shared between Dorset Museum and the British Museum in London. There has frequently been debate on whether the damage was caused by the Roman army but instead it is more widely believed that is was a domestic fight.
Conquest
Ancient Britain had become a political pawn in the eye of the Roman Emperors. Julius Caesar invaded in 56BC but it was a mere dip of the toe in the water. It wasn’t until a century later that emperor Claudius set his sights on the feral land. His entrance was impressive; arriving on elephants was a sure way to shock the local residents. However, it was Vespasian who ventured west to conquer the south of England.
The Romans arrived in Dorset at Poole Harbour, Radipole and the Fleet and moved inland, while other groups travelled west from Old Sarum and Winchester. Although struggles seem to have occurred, no major battle has been proved, however excavations at the hillfort of Crawford Castle (Spetisbury Rings) unearthed 130 skeletons suggesting the some of the Durotriges did fight back. As Vespasian gained ground he set up a number of forts in order to ensure future control and management.
Lake Farm Roman Fort
Hidden under the A31, the roundabout and floodplains of the River Stour is the mighty Roman Fort of Lake Farm. The area has been partially excavated on a number of occasions concluding that it is a site of considerable complexity. Evidence of earlier settlers, such as Neolithic and Iron Age pits, has been unearthed but it is the Roman period which dominates.
Vespasian, who was later to become emperor, constructed the military fortress, now known as either Lake Farm or Lake Fort, in order to ‘take on’ the native, but segregated, Durotriges Tribe. The location was crucial. It sat on the banks of the River Stour providing not only easy transportation but a good defensive barrier. The river also happened to be an important link for the Durotriges, Lake Farm then cutting off, or controlling, their coastal connections from the Iron Age Hillforts of Dudsbury and Hengistbury Head in the south to Badbury Rings, Buzbury Rings, Spetisbury Rings, Hambledon, Hod, Banbury and Sturminster Castle, to the north. The site also created a strong trade route to the salt marshes and sea at Poole Harbour.
Lake Farm matured to approximately 29 acres (11.7ha) in extent, creating a vexilation fortress. Internally, administrative buildings, centurion’s quarters, roads, rubbish pits, wells, buckles, pottery, tiles, hooks and fragments of armour have all been found while outside is evidence of industrial activity.
The most recent research, carried out by Bournemouth University and led by Miles Russell, has led to the identification of Vespasian’s own home. The most important building, or building for the most important person, is usually found within the centre of a Roman settlement. Geophysics has produced a plan which identifies an area that matches the theory. The location therefore sits right on the edge of the new A31 bypass, which itself replaced the 19th century Somerset and Dorset railway embankment in 1981. Therefore we may have missed the chance to excavate, leaving Vespasian’s home forever lying under passing wheels. Another aspect of the research suggests that Vespasian was living there with his wife. His daughter, Flavia Domitilla Minor, was born in AD45, a year after her father’s arrival, and possibly raised in the now lost Roman Fort. At the age of 15, she married Quintus Petillius Cerialis, who had just quashed the rebellion of the Iceni led by Queen Boudicia. She died in AD69 in Rome, possibly having only just arrived in the Roman city for the first time in her life; a true Dorset girl. After the Roman Centurions left the fort in the 60’s, it was abandoned. Medieval buildings and ovens have been discovered but the settlement gradually moved to what is now Wimborne Minster, no doubt still taking advantage of the Roman road network, the river and its fertile floodplains.
Wimborne Minster
Today Wimborne is dominated by its Saxon church with its two great towers built in the 12th century. Inside, on the south wall, is an astronomical clock thought to date back to the 14th century. The moon and the sun orbit the earth on its 24 hour dial, displaying Dorset time (8 mins or 1 hour 8 mins, depending on it being GMT or BST, behind London). The moon also displays its phases. The Minster is famed for its chained library, which is one of the few surviving chained libraries in the world. However, hidden between the pews is a floor tile that can be raised to reveal a Roman pavement, believed to be part of an old Roman temple. The discovery strongly suggests that this was an important religious site centuries before the Saxons claimed it as their own. Its presence may have also dictated the future movement of the population.
Shapwick Roman Fort
Shapwick (Vindocladia Burgus) Roman Fort features white chalk ramparts, rectangular buildings and three concentric ditches. It was undoubtedly part of the initial Roman advance through the area, probably established around 44 to 45AD, to provide protection for the construction parties working on the nearby fortress of Lake Farm.
Hod Hill
Excavations in the 1950s unearthed ballista bolts and sling stones suggesting the residents of the hillfort defended their home against the Roman invasion. However, their efforts were in vain as their chief was killed in battle before taking up his spear, which was found at the entrance of the largest hut circle. Under Vespasian, the Romans proceeded to use the ramparts in the north western corner to develop their own smaller rectangular fort, sitting on the highest point of the hill. Two gates were inserted into the new south and eastern Roman boundary with access down to the River Stour on its western edge. With views across the Blackmore Vale and down the Stour valley the base of 600 men and a 250 strong Calvary unit commanded the landscape until its abandonment in 51AD.
Wareham
It is strongly believed that the Romans set up a fort type settlement at Wareham. The two rivers of the Piddle and Frome would have been important trade and communication routes enabling the Romans to enforce control. Any evidence was trumped by the Saxons creating their own town so is buried under the present day layout
Waddon Hill Roman Fort
Waddon Hill was the last section of Dorset to be overpowered as the Roman soldiers marched their way across the chalk hills. It is possible to assume that this was their base as they ‘introduced’ themselves to the population of the surrounding hillforts of Coney’s Castle, Lambert’s Castle, Pilsdon Pen and Lewesdon. However this introduction could not have been that peaceful as two Roman ballista bolts were found on Pilsdon Pen. A plethora of other evidence has been unearthed too. Quarrying in the late 19th century on Waddon Hill uncovered Roman coins, Roman pottery, belt buckles and a first century Roman iron scabbard, all possible to see in Bridport Museum. In 1959 excavations took place to get a better understanding of the site. It resulted in being able to map out a full layout of the camp, apart from sections destroyed, and dates the site accurately to around 50-64AD. After only about five years, the fort was abandoned, this could have been due to a need for the troops in the east to combat the threat of the Boudican revolt, and/or the main route to Exeter and the west being moved closer to the coast. The Roman road to Waddon possibly left the main Dorchester-Exeter road at Eggardon Hill, to enter Waddon from the east. On an eposide of Time Team, Roman artefacts were found indicating a road travelling south from Picketts Farm that could be related.
There is no public right of way on the summit of Waddon Hill; however the Netherbury walk runs along its eastern edge and the Lewesdon walk along its western edge.
Beaminster museum plan an exhibition on the context of Waddon Hill Fort in 2025.
Black Down Roman Fort
The rectangular monument is enclosed by a bank and ditch with a small eastern entrance, marked by a pair of post holes possibly representing a gate, breaking up the bank. It was excavated in 1970 and was believed to be a Roman Fort constructed around the period of the conquest as a signal station. It also has extensive views towards the Dorset coastline to the south and in a potential line of site north-west to the Roman fort on Waddon Hill. The gated entrance which was removed suggests it was abandoned during the first century AD. However, a range of Bronze Age pottery has been found and, with the inclusion of the bank and ditch, it does not reflect the usual method of Roman construction. Other theories suggest it was associated with the large field systems which line the slopes to the south west, the earthwork then utilised later by the Romans.
Bradford Abbas
In 2010 crop marks to the north of the village revealed foundations of a Roman Camp, dated to around the 1st century AD. It is placed on a small spur with far reaching views to the North West into today’s Somerset.
Muckleford
On the slopes above the river Frome a military fort was discovered identifying the possible base for the engineers who constructed, what is believed to be, one of Roman Britain’s largest feats of civil engineering. The Roman aqueduct travelled through the landscape to feed the fountains, spas, baths and toilets for the Roman residents of the growing town of Durnovaria. It was abandoned in 150AD after approximately 50-70 years of use, possibly due to the failure of the dam that sealed a reservoir which fed the aqueduct.
Hamworthy (Moriconium)
On the edge of Poole Harbour, sheltered from the elements thrown from the sea, was the port of Moriconium. The Rivers Piddle and Frome provided access from the harbour to the west but the north relied on the roads, which provided an easy connection to the military fort of Lake Farm. In combination with the industries of stone, clay and salt occurring in the area it also created an easy trade route.
Radipole
During the Roman times, Radipole Lake was tidal and much larger. Roman evidence has been found to support the idea that the lake was not only being used as a reservoir but also as a shipping port, with the ability to sail ships right to its northern edge. Remains of a small landing stage have been found alongside evidence of settlement.
Norden
Excavations carried out at Corfe Castle have unearthed (from its deepest layer next to the chalk) Roman pottery. A theory has been put forward that a Roman fort or watchtower was constructed on the natural peak, which has since been trumped by the Norman castle. The location would have provided those watching from within with views over the workings of the Purbecks and towards the important harbour of Poole.
Church Hill
Above Whatcombe Wood, in the heart od Dorset, towering above the source of the River Piddle, is another rectangular enclosure which has been suggested to be a Roman fort. The interior is featurelss and has never been excavated but has a similar design to others in the county. It is also located in a strategic position, placed half way between Muckleford and Hod, guarding the ancient ridgeway that splits the Blackmore Vale from the south of Dorset.
Roman Roads
The roads were crucial to the expansion of the Roman Empire. They created strong communication and trade links but originally meant quick movement of troops at the time of the conquest. The roads were built to a high standard (survival rate alone supports this), combining a raised causeway surfaced with flints and flanked either side with ditches to help with drainage. It is clear that the Roman engineers had little respect for, what would have been the very impressive, spiritual monuments. Instead, to clearly secure their authority over the landscape, the Romans simply quashed them. The flints were gradually removed over time for local building but many of the raised earthworks are still visible today, some still used by modern traffic.
Ackling Dyke
Ackling Dyke connects Old Sarum to Badbury Rings and on to Dorchester (Dunovaria). As soon as it enters the county it trumps all existing earthworks. Breaking its way through Bokerley Ditch the route is still used by modern day traffic until it veers off, skimming the Bronze Age cemetery of Oakley Down. The road then cuts straight over the Dorset Cursus, a mighty earthwork that spanned 7 miles, before skimming the Iron Age settlement of Gussage Down. It continues in a south westerly directing to arrive at the hillfort of Badbury Rings.
Badbury Rings
Five Roman roads converge at the rings, creating a junction. The most famous road, Ackling Dyke, runs from the North East from Old Sarum (Sorviodunum), to the South West, to Durnovaria (Dorchester). Two roads travel north one to Aquae Sulis (Bath) and the other to the Roman Fort on Dunim (Hod Hill). A southern road travels to the fort at Lake Farm and on to Moriconium (Poole Harbour/Hamworthy). Sections of earthworks do occasionally rise from the ground around the hillfort while trees and fields also line the old route, leaving their marks as boundaries. The crossroads of Acking Dyke with the Bath road is unique as no other Roman roads cross anywhere else in the in the county.
Hod Hill and Bath
This road to Hod leaves Badbury Rings in a north westerly direction, crossing the River Tarrant at a ford before climbing up the hill towards Blandford. From here the route has been lost, under the modern day military camp, but would eventually have had to cross the River Stour. The road to Bath also fades as it leaves the rings but is clearly visible as occasional crop marks and farm tracks. The route is particularly significant as it passes a number of other archaeological sites, including Caesar’s Camp (an Iron Age hillfort) to the east and Roman villa to the west.
Hamworthy (Moriconium)
The road south heads to the fort at Lake Farm, crossing the River Stour on the way. Its raised earthworks are visible in the surrounding woodland at Pamphill but the ford has been lost to the water. A road diverts off to the east to Winchester via the New Forest. Its route through Dorset has also been lost but reappears in the landscape closer to the larger town. From Lake Farm the route then continues south to Hamworthy on the shores of Poole Harbour, cutting through the heathland of Corfe Mullen and Upton.
Weatherby
Ackling Dyke continues south west, past the Roman town of Vindocladia to cross the River Stour at Shapwick. It runs along the chalk hills, through the village of Winterbourne Kingston and then south of Weatherby Castle. Weatherby Castle, anciently known as Bindogladia, is an Iron Age hillfort built in the 3rd century BC, and, like Maiden Castle, evidence shows that the fort was abandoned in AD43, shortly after the Roman arrival. This particular section would have incorporated a ford as a crossing over the small river that runs from Milborne St Andrew to Bere Regis, now a small bridge. It may have also been an important junction at one time, connecting the east/west Roman road to the north/south, possibly older, route which leads to a villa in Dewlish in the north and the coast to the south.
Puddletown
Before arriving in Dorchester, Ackling Dyke runs though the heathland and forest of Puddletown, better known for the Victorian poet Thomas Hardy. He features the old road in many of his novels and poems calling it the Icen Way. The name is in reference to the Iceni Tribe who inhabited the east of England during the Roman invasion. The route continues straight though Puddletown Forest where a section of well-preserved remains were recently discovered. An image of a Roman solider has been reported on a number of occasions, hovering in the forest. It was later discovered that he was on the old Roman road of Ackling Dyke; his feet standing at the height the road would have been during the 1st century AD.
Weymouth
To the south of Dorchester the Roman road leads over the South Dorset ridgeway to the, then navigable, port of Radipole (Clavinium). From Dorchester the A354 uses its path, which was used for many centuries until quarrying began, diverting the road slightly west as it dropped down to Upwey. In 1824 the route from the peak was diverted slightly east, introducing a tight hair pin bend, to try and save the horses the struggle of the steep climb. The same road was used for the next two centuries until the development of the dual carriageway in preparation for the 2012 Olympics.
Long Ash Lane
To the north of Dorchester the Roman road follows the valley and aqueduct to Stratton, then splits in two. One runs through Maiden Newton in the direction of the Fosse Way while the second, known as Long Ash Lane, heads to Lindinis (Ilchester). The ridge top route has been in use for centuries, but passes through no settlements. Instead it skims the haunting sites of Neolithic long barrows and Bronze Age burial mounds and passes a collection of Roman Villa sites including Rampisham, Halstock, Frampton and Wynford Eagle. Thomas Hardy referred to its isolation in his most famous novel ‘Tess of the D’Urbervilles’ as Tess travelled alone across the landscape from Flintcombe Ash in the Blackmore Vale to Emminster (Beaminster).
Exeter (Isca Dumnoniorum)
The final road to leave Dorchester, other than the Roman bypass that skims the town to the north, is the Exeter (Isca) road. Its route is still used by traffic today on the busy A35 until the new road curves away from its older route near Winterborne Abbas. The Roman road then climbs up the hill is a zig zag fashion to Two Gates and on to cruise the ramparts of Eggardon hillfort. The road then disappears into the clay of the Marshwood Vale, where its path is debated. Investigations were carried out in the 1970s which concluded that a road possibly diverted north at Two Gates. Heading up along the ridge, avoiding the boggy and wild Marshwood Vale. It merges with the ancient Harrow Way to the north of the Beaminster Bowl near Horn Hill, then continues to the fort of Waddon. The route leaves the county having passed the hillforts of Lewesdon, Pilsdon and Lambert’s Castle. Another small road was suggested in an episode of Time Team, while researching the history of Pickett’s Farm, heading north to the Fosse Way.
More recent studies, carried out by Exeter University, also suggest the Roman road from Dorchester (Durnovaria) to Exeter (Isca) ran along the northern ramparts of Lambert’s Castle, circling the tricky conditions of the Marshwood Vale. Nevertheless, the route is still debated. Other theories still suggest that it continued directly west, with evidence discovered at Hogchester Farm in 2006 by Bill Putnam, unearthing a ditch and stone material, possibly hinting at a Roman drainage ditch and agger. Another possibility is the presence of two roads, maybe even similar to our modern systems of A and B roads. Today the road divides the counties of Dorset and Devon.
Diverting south west, away from Lamberts Castle and the possible northern road, is an impressive avenue of beech trees growing from raised medieval boundaries either side of an old wide track. The path must date back further with a direct route leading from the Iron Age Hillfort of Lamberts Castle to the coast. Its wide width also puts forward the theory that it may have been strengthened by the Romans, possible even connecting potential northern road, claimed by Exeter, to the southern evidence, found at Hogchester by Putman.
Roman Industry
Agriculture was the fundamental industry. Combining arable and pasture practices were encouraged by more efficient tools brought in by the Romans, as well as products such as carrots, turnips, peas and cabbage. They also introduced the cat and a selection of flowers like the rose and poppy, all centred on the new villa complex. However Dorset was also harvested for other materials.
Salt
Any accessible coastline may have been used during the roman period to extract salt. It was an essential ingredient for Roman life, used for preservation, seasoning, medicine and rituals. However it was an expensive commodity due as its difficult extraction methods. It was such an important aspect of Roman life that the soldiers were paid in salt and is the root of the word salary. Baked clay fragments found on the shoreline of Poole Harbour provides evidence of salt extraction works.
Clay and pottery
Pottery remains provide a great deal of information regarding life and trade during the Roman period. The best pottery was samian ware which was imported from abroad but the harder, coarser ware was manufactured in Dorset, known as Black Burnished Ware. It became one of the most popular makes, used for cooking pots, pie dishes, and flanged bowls. The army even took a contract out to supply the soldiers stationed at Hadrian’s Wall, with many fragments found on the walls course. Evidence suggests that it became a thriving industry on the Isle of Purbeck, transported through Poole Harbour, possibly being the centre of Britain’s pottery industry. A large pottery production site has been found on Arne which was featured in the TV series of BBC Two’s Digging for Britain. Many Roman kilns have been found around the county too with five found around Bradford Abbas.
Kimmeridge Shale
The black mudstone rock at Kimmeridge created the small cottage industry of producing Kimmeridge Shale. Prior to the Romans arrival it was used mainly for burning but it was soon discovered that it could be polished to create a high gloss finish. It was used mainly for artistic purposes such as decorative furniture, ornaments and jewellery, with many skeletons found wearing the products.
Stone
The shelly appearance of Purbeck marble, which could be polished to a high degree, attracted the Romans. The hard stone was used as grinding mortars and tablets for important inscriptions, such as those St Albans and the inscription from the temple of Neptune and Minerva at Chichester. Building would have increased sharply in the 1st to 2nd centuries increasing the demand for the stone exponentially. Houses, town walls, quoins and roof tiles were all quarried from the Purbecks, helping build some of Roman London. Portland stone was another valuable commodity and was used for many buildings and coffins by the 4th century. After the Romans departure, despite a slightly sleepy period, both Portland stone and Purbeck marble continued to be in high demand and were used for religious and important buildings right up to the present day.
Roman Towns
There are two main Roman towns (that we know of) in Dorset – Durnovaria and Vindocladia, both of which have developed in very different ways. A smaller town at Norden, on the Isle of Purbeck, has also been suggested. Outside of the county boundary are a number of other important towns including Isca Domonuium (Exeter) to the west, Lindinius (Ilchester) and Aqua Sulis (Bath) to the north, Sordininuim (Old Sarum) to the north east and Venta Belgarunm (Winchester) to the east. Like these cities, Dorchester has grown from the seeds of Roman settlement, saved from the beating of the plough. In contrast, Vindocladia has been razed to the ground.
Durnovaria
Aqueduct
Dorchester was first developed during the Roman period as a fort by the river during the siege of the nearby Iron Age Hillfort of Maiden Castle. When the castle was overthrown, the river settlement, founded in 60AD, grew into Durnovaria, a highly populated Roman town. This new ‘modern’ settlement included villas, mills, fountains, markets and public baths, all of which required a constant supply of water. One of the earliest installations was the aqueduct, believed to be one of Roman Britain’s largest feats of civil engineering. Built in the 1st century AD, it was used to transport fresh water from a reservoir, around 3 miles away, to the town. All along the valley are earthwork remains with some stretches well preserved; visible from the bypass, from Poundbury Hillfort and Bradford Peverell. Its path continues north west, to Southover, where remains of the Frampton villa were found. The amount it was used is still debatable but it was clearly a necessity as their impatience during its construction led to them using the natural channels of Compton Valence as a temporary feed.
The Roman reservoir was built by the engineers of the Second Augusta Legion between Muckleford and Frampton and connected to the Roman aqueduct through a dam. The development of the reservoir was an example of their outstanding engineering skills to ensure the town never ran dry. The 600m dam at the entrance was discovered in 1998, which would have held back, a capped, 270 million litres of water and served Durnovaria with 3,600 litres per minute.
The reservoir was abandoned in 150AD after approximately 50-70 years of use, possibly due to the failure of the dam, and the aqueduct was abandoned not long after in the 2nd-3rd centuries. Today little remains but only a few Roman aqueducts survive in the country, mainly focused around Hadrian’s Wall, making these earthworks all the more important.
Roman wall
The new town was protected by the construction of a wall. It surrounded the town on three sides with the fourth defended by a natural slope dropping down to the River Frome. The last remains of the wall have since merged into another wall and itself is protected from the outside world by iron railings. Despite the majority of the wall having been lost, its path determined the settlement’s pattern for centuries to come.
Roman Town House.
The town house was found when excavations occurred in the 1930s. A number of houses and other associated buildings were also discovered, but this was the best example. Complete with underfloor heating systems and floor mosaics, it is now on view for all to see.
Poundbury
The Roman cemetery at Poundbury is well known, sitting on the south western corner of the ramparts of Poundbury Hillfort. It was illegal during the Roman period to bury people within the town walls and therefore bodies are usually found on the outer boundary. The majority of burials date to the late Roman era of the 4th century AD, although the cemetery was in use from the Neolithic times to the Middle Ages. Some of the Roman burials were found decapitated with their heads by their feet! On display at the Dorset Museum are a selection of artefacts discovered within the Roman cemetery including a lead coffin, a child’s coffin, a sliver spoon and some plaited auburn hair.
Baths
Under the present day Wollaston Field Car Park, are the remains of the Dorchester Roman Baths. Excavation took place in 1977, when a large stone bath-house was revealed. Walls, water tanks, mosaics and hypocausts were also discovered serving several suites of hot, cold and tepid rooms and saunas. The presence of a number of pottery drinking vessels suggests that there were also rooms for feasting, banqueting and meetings all occurring within the baths. The substantial size and contents indicate that it could have been an important social centre of the town.
Over time much of the stone was reused in the buildings of Dorchester, but the remaining foundations were covered with sand and reburied to protect them after excavation. Today they still lie hidden under the car park.
Amphitheatre
Maumbury Rings dates back to the Neolithic period but was adapted for use by Vespasian’s army into an amphitheatre. The ground was lowered and flattened, to create an arena, and the extracted chalk was added to the surrounding banks to about the height they are today. The arena became one of the largest in Britain – but among the smallest in the Roman Empire, bordered by a wooden safety fence and capable of seating up to 10,000 people in seats carved into the banks. It would have been used for a number of functions from practice and demonstrations of army skills to wild beast shows and bull fights. Animals could have included wolves, bears, or wild boar, all of which were in Britain at the time. They may well have been gladiatorial fights (although there is no evidence of this) and quite likely a place of execution. Finds have included coins, pottery, bones and a burial but lead to it being abandoned in the 4th century.
Vindocladia
Vindocladia was placed in an area where a number of Roman roads cut across the landscape. The name derives from the Celtic British to ‘windos’ meaning white and ‘klādiyos’ meaning ditch, translating to ‘The Place of White Walls’, this could be in reference to a defensive boundary constructed around the Roman town from the abundant chalk lying directly underfoot. Evidence discovered supports the theory that the settlement was active during the late Iron Age and then developed throughout the Roman period to the 5th century. Artefacts unearthed have included glass beads, bronze pins, bracelets, brooches, Durotrigian silver coins, Roman coins, stone roofing slabs, painted wall-plaster and pottery. The suggestion has also been put forward that it was one of the largest Roman towns in the county, second only to Durnovaria.
As the ruins of the once great town crumbled, the land was farmed, with the name Shapwick deriving from the Saxon ‘sceap’ meaning sheep and ‘wic’ meaning village. The old Roman ford through the River Stour also began to breakdown, trumped by Crawford Bridge to the north and White Mill Bridge to the south, both amongst the oldest river crossings in the county.
Amphitheatre
Alongside the famous amphitheatre of Maumbury Rings is possibly a second at Vindocladia. Excavations have taken place on the site which unearthed a pregnant young sow dating back to 1100-900 BC, possibly related to animal sacrifices and suggesting that the amphitheatre was built upon an older earthwork. Over time is has been majorly ploughed out, but, since its discovery, the farmer has let it grow, protecting it from any further damage.
Norden
In 1984 excavations were carried out at Norden, north of Corfe Castle. A large number of historical elements were found, relating to both industry and settlement, including pottery, gravel surfaces and oyster shells. The suggested Roman town was placed in the heartland of the industrial Isle of Purbeck with three villas (East Creech, Brenscombe and Bucknowle) growing nearby. Not only may it have become the administrative hub of the area it could have also guarded the gap through the chalk ridge, on which Corfe Castle now sits. The modern village of Corfe has since moved to the south of the gap, where the entrance to the Norman castle was placed.
Roman Villas
The definition of a Villa is debated. They are usually farmsteads that have developed during the Roman rule into a courtyard building with a collection of rooms off a corridor, decorated with floor mosaics and heated by hypocausts. Latin for farm was also simply ‘villa’. Villas often had slaves too and slave babies were not permitted. The total of 97 baby skeletons were found outside the villa at Hambledon in Buckinghamshire suggesting they had a zero tolerance rule. Within Dorset, Villa distribution is uneven, with little explanation, but not all have been found! However, their absence may prove to be more informative. Cranborne Chase is one example. Despite the plethora of ancient activity that scatters the chalkland landscape, its heart is villa free; possibly suggesting it was not economic enough for industry, too harsh to farm, too feral for settlement or too peaceful to guard.
Halstock
The Halstock villa was discovered in 1818 by road workers who had accidently unearthed an intricate mosaic. Many villagers saw this as an exciting find and so proceeded to visit the site and depart with a number of their own souvenirs. The Earl of Ilchester, the owner of the land at the time, decided to recover the villa for its own protection.
Between 1967 and 1985 the site was excavated a number of times by Ron Lucas. It was concluded that the villa had started life as an Iron Age farmstead, accessible from the neighbouring ancient Harrow Way. The site was then developed during the Roman occupation and became a large complex which included a number of buildings and courtyards crossing the old lane. The buildings were constructed to include hypocausts and a range of mosaics suggesting it was a villa of some wealth. There are no Roman roads nearby, possibly suggesting that the Harrow Way was efficient enough. A number of artefacts have been found on the site including a collection of coins dating between AD260 to AD375. The area was turned into a golf course, which has since closed.
Rampisham
In 1799, a well preserved Roman mosaic, measuring 4m by 3m, was found on the slopes looking over Evershot. A pattern of rings and a floral decoration filled the design, but it was destroyed by treasure-hunters. However, it indicates that a large, impressive villa must have been present here at one time.
Wynford Eagle
At first Roman coins were unearthed in the river valley, covering the reigns of Nerva, Trajan and Hadrian. In 1935 a mosaic pavement with a dolphin and flowered borders was uncovered which has led to it being confirmed as a Roman villa site.
Frampton
The village of Frampton has a rich history, highly influenced by its proximity to the Roman town of Durnovaria (Dorchester). In the valley is Nunnery Mead, a nature reserve consisting of a thin strip of woodland on the southern banks of the River Frome. In the late 18th century a Roman villa was discovered here by accident when men were digging for flints for construction purposes. It was first excavated in 1796 and found to contain impressive mosaics decorating large rooms connected by corridors as well as a shale table leg (most probably made in Kimmeridge).
The majority of the evidence unearthed was dated to around 350AD. In 1797, George III inspected the mosaics, most probably a trip out from his frequent visits to Weymouth. In 1963 further excavations were carried out, but little was found other than a wall, a few vases, coins, pottery and a stone column. The majority of this evidence still lies beneath the surface, having been reburied. The rest are in the form of earthworks.
Lenthay
A mosaic depicting Apollo and Marsyas was discovered in 1836. It was found in a large field to the west of Sherborne. It is believed to have come from a wealthy villa, placed in the fertile valley of the River Yeo.
Thornford
As the River Yeo meanders its way towards Bradford Abbas, neighbouring its banks are the remains of a Roman villa. After the discovery of a number of items, which included pottery and tesserae during drainage works, the villa was excavated by Professor J Buckman in 1876. It was excavated for a second time in 1961 by Mr J Leach and the Sherborne School Archaeological Society. A large complex was unearthed consisting of a six rooms, four of which were decorated with intricate mosaics and connected via a system of corridors. A large number of animal bones, pottery and coins, dating from AD211 to AD370, were also found with a lot more still left to be unearthed.
Charminster
Roman roofing tile and Romano British pottery have been found in the valley, while on the opposite slopes of the Cerne valley a tessellated pavement was discovered in 1891. In 1960 further investigations were carried out unearthing window glass, decorated plaster and hypocaust tiles possibly marking the site of a major Roman villa.
Fifehead Neville
On the eastern banks of the River Divelish, is the site of what was once was of Dorset’s most elaborate and impressive buildings. In 1880 and 1903 the remains of two extensive wings of a Roman villa were found, complete with columns, floor mosaics, hypocaust systems, lead piping, a rectangular plunge bath, a horde of tools and early Christian jewellery. The jewellery included two silver rings found engraved with the Chi-Rho symbol, an ancient Christian symbol, suggesting that they were wealthy Christians that lived here. Many of the artefacts are on display at Dorset Museum. It was placed near the Roman road which connected Hod Hill to the bustling Roman town of Durnovaria, and was just south of another Roman building discovered in Hinton St Mary.
Iwerne Minster
The area of Park Farm, on the banks of the Fontmell Brook, has produced evidence of Iron Age and Roman occupation. It was excavated in 1897 by General Pitt-Rivers, it being his last dig, passing away before its completion. The Iron Age finds included Durotrigian silver coins and a bronze brooch. In early Roman times the settlement was modified and finds of this period included coins from Vespasian (AD69-79) to Commodus (AD176-192), brooches, and samian pottery of the 1st and 2nd centuries. Evidence also gathered supports the theory that in the 3rd century a large building, nearly rectangular in plan, was constructed that included a cattle shed and a granary as well as decorative plaster work. Coins ranging in date from Gordian I (AD238), Maximian (AD283-305) and Decentius (AD350-353), a bronze belt-link and a bone weaving comb were all found in or near the building.
Tarrant Hinton
In an empty field between Tarrant Gunville and Tarrant Hinton near the River Tarrant, a range of Roman activity, covering 20 acres, was discovered in 1845 while excavating chalk. Remains of foundations and painted walls, outlining rooms and corridors, had floors paved with red and white mosaic tiles. A 30ft well was found containing pottery, pipes, rings, brooches and coins, dating from Constantine and Constantius of the 3rd century AD. All the evidence strongly pointed to an elaborate and wealthy Roman villa.
The landscape is also scattered with Roman remains. To the north is Caesar’s Camp, a possible Roman fort. Across the valley to the east is the site of a Romano-British settlement and the villa at Farnham while further to the west is the vila at Iwerne Minster and the Roman fort of Hod Hill. The landscape is split both sides by the Roman Roads from Bath and Old Sarum shooting through on their way to Badbury Rings.
Farnham
In 1868 a villa just outside Farnham, on Oak Hill, was discovered. Foundations of a building were unearthed alongside oblong and diamond shaped roofing tiles. Painted wall plaster with coloured borders of red and green were alongside red and white tesserae, all in the company of scattered coins. Little else is known about the site and has since been reburied.
Mycern Farm
The Mycern Farm Roman Villa is impressive. At over 35 meters long it is in the company of a number of other buildings including a bath house, a multi-story barn and a structure that may well have been an outdoor pool. There has also been evidence produced to suggest industry in cloth making and metal working alongside agriculture and hunting. Internally the villa was decorated with high quality mosaics (possibly from Ilchester), painted wall plater and columns.
Hemsworth
The Hemsworth villa was first discovered in 1831 but it was not fully uncovered again until 1908 and much of it had been lost to the plough. It sits on a small spur between two Roman roads heading for Badbury Rings. Excavations have unearthed a number of wings, connected to a main block via a corridor, some with hypocausts and mosaics. The largest is centred on a sea god and another around the goddess Venus, while the smallest, with black and white colouring, was used as the floor of a plunge bath. Coins and columns have also been found but there are many signs of burning suggesting the villa was destroyed by fire. A small section of a floor mosaic is up in Dorset Museum. It depicts the sea god Neptune, with crab legs and claws growing out of his head. A larger section is in the British Museum.
Preston
A Roman villa was discovered in the fields between the town and sea in 1844. Around the same time the discovery of the Roman Temple took place which overshadowed the villa. No external elements remain but excavations in 1852 and 1871 unearthed mosaics and pottery dating from the 3rd to 4th centuries. Further investigations took place in 1932 concluding that it was a much larger building than first thought with more industrial activity occurring. A bath house and a furnace was discovered along with coins of Corausius (AD286-93) and Constans (337-350), roofing tiles, tesserae, limestone slabs, a quern, Samian and New Forest Ware pottery, painted wall plaster and a Kimmeridge Shale furniture leg. In 1946 the mosaic, having become damaged from frequent visitors, was sealed with a layer of concrete and reburied.
Nearby Roman Bridge Cottage dates back to the 17th century. Just within its gardens is the little bridge that gave the house its name. The ‘Roman Bridge’, now buried under brambles and bushes, is a little, single arched bridge believed to have been built in the early 18th century with 16th century foundations, but its exact date is uncertain. Hearsay and folklore claims that it dates back much further, to the Romans, therefore connecting both the temple and the villa to the Roman road to Dorchester. However analysis of the construction has suggested a Norman origin. It would have at one time been an important crossing point over the River Jordan but was finally trumped by the turnpike and then newer bridge to the south in the 19th century. Its future may be in jeopardy as its site is slowly forgotten. In 2006, at Weymouth Bay Caravan Park to the south, a second bridge was discovered. An area that was densely covered in vegetation was cleared revealing what could have been part of a Roman road connecting the villa to the river crossing and the Roman temple. Little investigation has occurred since.
Druce Farm
On the slopes to the west of Waterston a Roman Villa was discovered. A series of studies were carried out between 2010 and 2015 concluding the villa had been a large complex, inhabited throughout the Roman period and beyond. It proved to be a typical villa with a winged corridor and a courtyard. However, once it was abandoned and fallen into ruins, the stone was used in the building of both Waterston Manor and the church. Nevertheless, a number of mosaics were unearthed, one particularly well preserved and decorated with a swastika, which was a symbol of good luck in the Roman period. Other artefacts found included pottery, glass, iron and even a silver ring. Like Mycern Farm evidence has been produced to suggest industry in cloth making and metal working alongside agriculture and hunting, while the mosaics also seem to originate from Ilchester.
Dewlish
Dewlish House was built in 1702 for Thomas Skinner and is set in a large 296-acre (120 ha) estate. In 1740, a storm uprooted a tree in the grounds and a Roman villa was revealed. Sadly, as their approach to archaeological recoding in the 18th century was less formal, many of the records have been lost. In 1974, during a more modern approach to archaeological investigation, evidence was found to support occupation from the Iron Age to the late Roman. The villa had developed over time into a complex of 24 rooms, including a bath suite, some heated by hypocaust systems. An elaborate mosaic was discovered depicting a leopard attacking an antelope. It was given to the owner of Dewlish House as a thank you for letting the excavation occur (fewer restrictions were put on these treasures in the 70’s). It was later auctioned off and bought by an international buyer. However on 17th July 2020 the export was suspended by Culture Minister Caroline Dinenage and £150,000 was raised to purchase it for Dorset Museum where it is now displayed.
Bucknowle
Bucknowle, which sits to the east of Church Knowle, has produced evidence of much earlier activity. Iron Age elements have been unearthed but what dominates its history is a Roman villa. It is the only Roman villa to be found in Dorset (so far) south of the Purbeck Hills and has proved to be a substantial development. Excavations carried out between 1975 and 1996 concluded that an aisled hall and bath house were first constructed in the 1st century AD. It slowly grew and mosaics, tiles and limestone slabs were found along with evidence for a large, three storey building with underfloor heating. The buildings merged and created an impressive Romano British complex until it was abandoned in 410AD. Today Bucknowle consists of a woodland cottage, built in 1863, and an impressive Portland stone house, built in 1870, for the estate and estate manager respectively of nearby Encombe. Bucknowle was eventually built on the site of the villa nevertheless many artefacts are still recovered, including coins, roof and flue tiles, jewellery and pottery. Painted plaster from the walls are on display in Dorset Museum along with a delicate glass jug.
Brenscombe
While the farmer was digging post holes for a new fence he exposed a Roman mosaic. Excavations were carried out in 1961 and 1967 when two mosaic pavements found. Other elements unearthed included brick, roof and flue tile, some pottery and a Kimmeridge shale disc. This has put the theory forward that the villa may have been involved in the shale workings at Kimmeridge. Coins from the 2nd and 3rd centuries have also been found along with a mysterious 3inch metal crucible.
East Creech
In 1869, during the usual clay extraction, a Roman villa, with mosaic floors and columns, was discovered between East Creech and Cotness. A few years later, in 1888, a four-feet high Tuscan column, a hypocaust and a red and white mosaic floor was uncovered. In the early 20th century evidence was discovers of a nearby Roman armoury, utilising local Kimmeridge Shale. This landscape is so organic that it is changing frequently due to its soft sandy terrain. It is therefore no surprise that although Roman discoveries have been made, much evidence has disappeared, a huge contrast to the chalkland environments elsewhere.
Shillingstone
In 2004, archaeologists from AC Archaeology uncovered the remains of a significant Roman villa on the outskirts of Shillingstone, during a housing development by Wyatt Homes. The site revealed a large two-storey building complete with thick walls, a bathhouse, mosaics, Roman coins, and local pottery, suggesting it belonged to a wealthy Roman family and served as the centre of a major estate in the third or fourth century AD. The team also found traces of plastered walls, hypocaust heating, and environmental evidence like burnt wood and snail remains, offering insights into the ancient landscape.
Charlton barrow
Evidence of a possible Roman villa has been unearthed, near Charlton Barrow, to the south of Charlton Marshall. Artefacts include pottery and elaborate jewellery suggesting a villa of some wealth. On the opposite side of the river, towards Badbury Rings, sat the Roman town of Vindocladia, now nothing but fields, while two Roman roads skim the parish, one traveling north west to Hod Hill and the second south west to Dorchester.
Winterborne Kingston
For over fifteen years, Bournemouth University students and staff, as part of the Durotriges Project, have been excavating an Iron Age to Romano British settlement at Winterborne Kingston. This site, which sits just to the north of the Roman road, includes a Roman villa and cemetery, providing rare and unique insights into the lives of its occupants. Excavations uncovered burials of two women, aged in their thirties, two sacrificed hunting dogs and a burial of a man in a coffin, all exhibiting a blend of Roman and local traditions. They are also believed to be the likely occupants, which is rare in Roman villa excavations in Britain. More burials were discovered during excavations in 2025.
Bagber
To the west of the Milborne Brook, high above the lost village of Bagber, a number of buildings and occupation debris, dating from the Romano-British period, were unearthed in the 19th century. A rectangular building was discovered, with pottery, 2nd-century coins and burnt stone, possibly suggesting the location of a Roman villa with a hypocaust. The range of pottery fragments also put forward the theory that it was a small industrial site.
Piddletrenthide
The18th century Piddletrenthide Manor House sits on the heart of the village. The house replaced an older building with some of the original stone work incorporated into the new design. In 1740 some artefacts were unearthed from the gardens believed to be mosaic from a Roman villa. Its location has been claimed to be suitable for a villa, being close to the River Piddle and at the western end of the valley to Plush. It is also in close proximity to the villas at Bagber, Dewlish, Druce Farm and Charminster. Romano-British debris, including roof tiles and pottery, has also been discovered on the slopes to the east, near Doles Ash Farm***, dating from the 3rd to 4th centuries. No further investigation has taken place on either site.
Roman Temples
The combination of Iron Age and Roman influence led to a mixture of beliefs, from Celtic gods to Roman deities. However, many of these gods shared attributes, for example the Celtic god, Cocidius, was similar to the Roman god of Mars, Maponus similar to Apollo and the water goddess, Sulis, similar to Minerva. In the 1st century AD, circa 63AD, tin merchant and uncle of Jesus, Joseph of Arimathea, had captured Jesus’s blood in a cup (used by Jesus in the last supper) known as the ‘Holy Grail’. He then brought it to Britain, arriving at Glastonbury, and Christian worship spread across the country in the 2nd century. The first bishops are recorded in the 4th century and at around the same time, Christian symbols appear more frequently in wall plaster and mosaics.
Maiden Castle
Within a few decades of the arrival of the Romans, Maiden Castle was abandoned. However, in the late 4th century, two temples were built on the hill. At this time, a fusion of native British and classical Roman religion was becoming popular, and it is common to find shrines located in remote rural locations. The abandoned hillfort provided an ideal setting for this new pagan religion.
The larger square building, at about 40ft, was built sometime after 367AD but continued to be updated over time, with new wall plaster, until falling into decay during the 5th century. The second circular building jut to the south of the square structure is about half the size of the larger temple but built about 5 years earlier. It is placed within the boundaries of an older roundhouse that sat at the head of the main street which ran through the fort, suggesting it was a building of some status. Excavations have unearthed a number of coins dating from the 4th century to the early 5th. Figurines have also been found on the site dating from this period, including one of a bull and another of the goddess Minerva.
Jordan Hill
The Roman Temple on Jordan Hill is currently managed by English Heritage. Its position offers wide open views over the Jurassic Coastline and inland to the Osmington White Horse. The temple dates to the end of the Roman period when it was beginning to fragment. The remains, the foundations visible, suggest a small temple with a central sanctuary, or cella, rising above a low columned portico which surrounded it on all four sides, similar to the temple discovered on Maiden Castle. A pit was discovered which contained coins of the Emperor Theodosius I (AD 379–95), old roofing slabs, two urns, a sword and a spearhead, all possibly related to ritual offerings. A hoard of over 4,000 bronze coins was also unearthed in 1928.
Between the temple and the villa a cemetery was discovered thought to date to the 1st century AD and in use to the 4th century, so throughout the Roman occupation. Only a small section was excavated which produced over 80 skeletons, some buried in wooden coffins (only the nails remaining) others lined with stone slabs. Personal objects were buried with them including pots, combs, jewellery, arrowheads and iron swords. The method of the burials and the type of offerings that were placed with the bodies also reflected the gradual Roman influence over the Iron Age tribes. Some of the finds are now in the Dorset Museum and the British Museum.
Hinton St Mary
The village is most famous for the discovery of an impressive Roman mosaic. In 1963, while digging a hole for his wife’s washing line, the village Blacksmith, Walter John White, discovered coloured tiles. The area was subsequently excavated with further excavations taking place as recently as 2022 and 2023. At first it was concluded that the site must have been a villa, with the unearthed mosaic depicting what is thought to be Christ. If so, it becomes the first known, and therefore oldest, depiction of the religious figure anywhere in the Roman Empire! Later excavations have proved that the site was much larger than thought, producing further mosaics. A new theory was put forward suggesting that, as the main room was standing alone, it could not have been a villa but instead a religious complex, decorated to the highest standards. The mosaic has been identified to correspond with the workmanship of other mosaics found in the area including at Frampton, Fifehead Neville and Hemsworth, and therefore all attributed to the mosaic workshop of Durnovaria. The mosaic found a home in the British Museum but, in 1997, despite huge outcry, it was broken up to make way for museum improvements. Today only the central figure of Christ is on display, the rest in boxes in archives. There have been many calls to return it to Dorset and negotiations between the British Museum and Dorset Museum are ongoing.
Duncliffe
Duncliffe’s history, although elusive, dates back to the Iron Age with the discovery of a bronze statuette of a boar, known as the Motcombe Boar. A second figurine of Fortuna, the Roman goddess of Fortune has also been discovered, thought to date to the 2nd Century AD. Theories put forward for their presence is that the hill was used for rituals with the Roman figurine possibly suggesting the presence of a Roman temple. Although no evidence for a structure has emerged so far the site location is similar to the Roman Temple on Maiden Castle.
Fordington
This Roman history and influence leaked through the walls that once surrounded Durnovaria into Fordington and Roman coins are frequently found. Bodies were found in 1810 just outside of the original Roman wall. They were buried with small glass and amber beads, buckles, bone rings and brass jewellery, while a female skeleton still wore a bracelet made from Kimmeridge shale. They are believed to have been part of a Romano-British burial-ground of Dorchester as burials were not allowed to occur within the town walls. It has clearly been a spiritual spot for centuries and prior to the construction of the church of St George an earlier church stood on the site, built in 857AD, and prior to that it has been suggested it was the site of a Roman temple. During construction of the new London road further Roman remains were found which included a hypocaust. A beaded necklace was also found at Fordington, now on display at Dorset Museum.
Norden
In the fields north of the National Trust Visitor centre in Corfe Castle, geophysical results returned to indicate the presence of a Roman temple. This would have placed it near to the Roman town of Norden, in the heart of the industrial Purbecks, and neighbouring three villas (East Creech, Brenscombe and Bucknowle).
